This study was initiated to test the accuracy of our statistical biochemical laborary approach toward identification of alcoholism. The test group consisted of patients seen in consultation on the medical wards for drinking problems as well as in self-acknowledged alcoholics seeking help in an outpatient alcoholism clinic (n=200). A control group of both inpatient (n=100) and outpatient nonalcoholics (n=100) as well as small prospective groups (n=40) of the two alcoholic and the two nonalcoholic groups were to be obtained but has not been forthcoming as of this date. Analysis of this data, on the basis of discriminant functions developed from data at the Long Beach, VA, could be used to identify the collected alcoholic patients as alcoholic, but this would not be ideal. Accordingly, a special effort is being made to obtain a control sample at Walter Reed Army Medical Center.